( 292 ) 



(see also Teyler 1905: Qnelques remarqnes sui- I'efination d'ètat, p. 46) : 



/•-// ^ ^ 1—6' 2 



But the values of h' and //' resulting from this will of course be 

 slightly different from those which we shall now calculate from the 

 same experimental data ; and so we are not allowed to draw any 

 conclusions from the deviation of the values of />' and //' calculated 

 by us just now from those found before. They remain simply secon- 

 dary values, which can never be found by direct ex|)eriment, but 

 only by the calculation based on certain suppositions. 



Now in the first place the value of h' can be found from (3), 

 viz. h:^h,, — (1 — 13) ^^>, from which directly follows: 



an 



h' = =1 Lh 



d V dv 



Ab 



Taking (d) and (1 -]- ./;,i) : = 'f mto account, we get 



V — b 



// = ^ (II) 



1 ^ -i3(i-,i){.v+cpr 



The calculation of //' from tliis expression is about as elaborate as 



d'p dp f V]. \ 



that of — - from — see § 2 the calculation of . The easiest 



dv"- uv \ vie — bi;J 



WAY to reach the result is the following. The numeratoi' of the 



(p f \ 



second member of (IJ) being = y we put // tor ^{l-^i){x^'f)^ 



.f -[-'/'' V * ■'•-f-1 



for shortness' sake, above in ^ 2 ), we have: 



hence 



.V-i-'f' 



du if^ dy ,v d'fj 



dv a-\-(pdv {•^^-T-ffY do 



But as 6 = = , we iiave 



X -~ rp X -~<p ^v -\- (fi I -}- y X -(-7' 1 + .'/ 



also 



ffj 1 du xil dp 



i'—f') fy (i+.y) = -^ Tx- ^'-^^ T + nf r («-^'') T- 



